Credit Risk Measurement Based On Macroeconomic Variables: Credit Risk Stress Test Application in the Banking Sectors of Turkey

The purpose of this study is to predict the ratio of non-performing loans as well as expected loss, unexpected loss and credit at risk values in the banking sector of Turkey as a result of macro-economic shocks under hypothetical scenarios. Credit risk models were developed by including macro-economic variables affecting the credit risk based on the Credit Portfolio View approach proposed by Thomas Wilson in 1997. Scenario analysis was employed as well as Monte Carlo Simulations with respect to significant variables in the credit risk models. In the credit risk models, non-performing loans was used as the dependent variable, while unemployment rate, interest rate, money supply, inflation and GDP were used as explanatory variables. When stress test was applied to macro-economic variables which affect the banking sectors in Turkey, it was observed that the shocks seemed to be influential on the non-performing loans